Peer-reviewed Research Article
Modeling of spatial variables using neural networks
Main Article Content
The paper examines a modeling of spatial variables using neural networks. The example of creating a one-dimensional mathematical model of the spatial variable of the geochemical field field shows the possibility of dividing the variability of the geological indicator into regular and arbitrary components using artificial neural networks. A method for selecting a neural network model suitable for a certain criterion for a simple component in a limited data set is proposed. The statistical analysis of the obtained results is given by building a regression model of the variance of the random component of the spatial variable.